Winboard Forum

I would like to know what to do with Protector :

http://sourceforge.net/projects/protector/

Do our experts consider Protector as a Fruit/Toga derivate ?

Olivier

Olivier Deville wrote:I would like to know what to do with Protector :

http://sourceforge.net/projects/protector/

Do our experts consider Protector as a Fruit/Toga derivate ?

Olivier

Without claiming to be an expert I'd say it's a derivate of some engines. The README says:

Protector is based on many great ideas from the following people: Fabien Letouzey (pvnodes, blending of opening and endgame values, eval params), Thomas Gaksch (pvnode extensions, extended futility pruning, space attack eval), Robert Hyatt (consistent hashtable entries), Stefan Meyer-Kahlen (UCI), Gerd Isenberg and Lasse Hansen (magic bitboards), Marco Costabla, Tord Romstad (Stockfish mobility count and more) and Mike Donnig (testing). Without their contributions Protector would not be what it is. Thank you so much.

However, the author hasn't forgotten to document that and is also open for a substantiated discussion.

vp

By that standard every engine in existence would be a derivative, with the exception of Pos 1.19 and NEG 0.3d...

H.G.Muller wrote:By that standard every engine in existence would be a derivative, with the exception of Pos 1.19 and NEG 0.3d...

I was about to post the same answer : implementing ideas is allowed, copying code is forbidden...

So the question is : did the author copy code from Toga/Fruit or other open source programs ? If not, Protector should be allowed to play.

Any comments are welcome.

Olivier

H.G.Muller wrote:By that standard every engine in existence would be a derivative, ...

And your suggestion to answer Olivier's question is?

Volker Pittlik wrote:

H.G.Muller wrote:By that standard every engine in existence would be a derivative, ...

And your suggestion to answer Olivier's question is?

Hi Volker/Olivier,

Leo asked me for my opinion about Protector. This was my reply (please bear in mind I don't claim to be an expert in clone/code analysis)

There is no 'copy & paste' code that I can see, but eval terms look Toga/Fruit based with some ideas taken from Crafty, Stockfish.

Jim.

I've looked through Protector a couple of times and I can't find any copied code. The similarity to Toga/Fruit is in the *organization* of the search code. But most of the details are different. The Toga/Fruit search method is extremely strong and could very well be the best way to search. I'm doing a total rewrite of Horizon and, when I get to the search, I intend using Toga/Fruit as a model for the search. Just like Raymond I will implement it in my own way, but the organization will resemble Toga/Fruit. Well, not quite -- I have an idea about something -- but there will be a similarity.

There is also this old thread that discusses the topic a bit.
viewtopic.php?f=2&t=50337

The unfortunate part about the Protector clone/derivative accusation is that the author brought it on himself by what he said it in his original readme file. I understood what he was trying to say, but that's not how most people interpreted his comment. He was trying to say that the algorithms used in Protector were not his invention and are well-known chess programming methods. He should have added that all implementations of the algorithms were his own.

Ron

Volker Pittlik wrote:

H.G.Muller wrote:By that standard every engine in existence would be a derivative, ...

And your suggestion to answer Olivier's question is?

I have no answer. I have never seen the code of this engine, nor have I ever seen the code of any of the engines by which, according to the README you posted, the author has been inspired.

But the README simply describes what everyone who developed a 2500+ engine in less than 15 years _must_ have done: make a thorough study of the open-source engine codes that are around, to see how it is done, and then do it themselves in different combinations, with some modifications.

Ron Murawski wrote:The unfortunate part about the Protector clone/derivative accusation is that the author brought it on himself by what he said it in his original readme file. I understood what he was trying to say, but that's not how most people interpreted his comment. He was trying to say that the algorithms used in Protector were not his invention and are well-known chess programming methods. He should have added that all implementations of the algorithms were his own.

Ron

I agree with Ron. I'm happy to continue testing Protector.
In addition, the author, Raimund, seems to be a nice enough and genuine fellow.

H.G.Muller wrote:
I have no answer. I have never seen the code of this engine, nor have I ever seen the code of any of the engines by which, according to the README you posted, the author has been inspired.

But the README simply describes what everyone who developed a 2500+ engine in less than 15 years _must_ have done: make a thorough study of the open-source engine codes that are around, to see how it is done, and then do it themselves in different combinations, with some modifications.

Hello H.G.,

I have no way of refuting this model of latter day chess engine development. In fact, I yield to you in matters like this.

My point is that I would be willing to bet a small sum that back in the days when I first came to this fascinating computerchess hobby there were tons more ideas and differences.

I may not know about code but I played against various engines and watched them play and these engines had personalities - as strange as that may sound.

Yace was different from Crafty which was different from Phalanx which was different from the Baron and that was a good thing.

Now the future appears depressing - a line of open and closed source engines inspired by a limited common ancestry.

I have played (and continue to do so incidentally) against your engine, Ursurpator, and it is not a replica of a 3000+ engine told to play like a 1500 player. It plays genuinely enjoyable chess. Well at least to me. It is different and I love it for that reason.

Perhaps that really is the future you are outlining and maybe you are correct but I genuinely hope not.

Later.

H.G.Muller wrote:

Volker Pittlik wrote:

H.G.Muller wrote:By that standard every engine in existence would be a derivative, ...

And your suggestion to answer Olivier's question is?

I have no answer. I have never seen the code of this engine, nor have I ever seen the code of any of the engines by which, according to the README you posted, the author has been inspired.

But the README simply describes what everyone who developed a 2500+ engine in less than 15 years _must_ have done: make a thorough study of the open-source engine codes that are around, to see how it is done, and then do it themselves in different combinations, with some modifications.

Not everyone...

Miguel